The candidate is currently a Research Associate in the Sidney Kimmel Cancer Center at Johns Hopkins who has previously spearheaded efforts towards discovery of new pancreatic cancer markers and the identification of invasion-specific genes in pancreatic cancers in the laboratory of Dr. Scott Kern at Johns Hopkins. The long-term career objectives of the candidate are to develop early detection markers and identify therapeutic targets for human melanoma. This research proposal uses both a mentored research training period (years 1-3) and an independent training period (years 4-5) in order for the investigator to develop gene discovery systems for human melanoma under the guidance of Drs. Rhoda Alani and Scott Kern. The investigator already has early data on promising new diagnostic/prognostic melanoma tumor markers (the HLH transcription factor, Id1 and in-situ telomere length) and therapeutic targets (Id1, telomerase) in melanoma and those will be developed as part of this proposal. The specific aims are, therefore: 1) to identify novel markers and therapeutic targets for melanoma using gene profiling of human melanoma cell lines and primary human melanomas from varying stages of disease progression in combination with experiment-oriented data reduction algorism; 2) to determine the diagnostic/prognostic significance of Id1 expression in melanocytic lesions and its utility as an indicator of malignancy and therapeutic target; 3) To evaluate in-situ telomere lengths in archival samples of melanomas and determine the diagnostic/prognostic utility of telomere length in predicting biologic outcomes of melanomas in a retrospective study. The overall goal of the investigator is to transition from research in the molecular characterization of pancreatic cancers to analogous studies in human melanoma. Since melanoma is the fastest growing malignancy in the U.S. and there are no therapies that have been shown to increase life expectancy for patients with advanced disease, it is anticipated that these studies will be of high impact. [unreadable] [unreadable] [unreadable]